Completion of Japanese Sentences
by Inferring Function Words from Content Words
Koji KAKIGAHARA and Teruaki AIZAWA
ATR Interpreting Telephony Research Laboratories
Twin 21 Bldg. MID Tower, 2-1-61 Shiromi,
Higashi-ku, Osaka 540 Japan
Abstract
A method of generating a Japanese sentence by
inferring function words from content words using
valency pa~terns is presented.
A procedure for selecting an appropriate function
word, on the assumption that correct content words have
been selected for a given phrase lattice, is described. A
method ol~ inferring a correct verb when verbs are
recognized less accurately than nouns by the speech
recognition system is described. Sentences are produced
from content words as inputs by using the valency
patterns ohtained from collected dialogue sentences in a
restricted 1ask domain. Using the semantic features of
preceding nouns and valency patterns allow a [airly
develop an interpretation system between Japanese and
English has already started. For this purpose, however, an
effective interface between speech recognition and
machine translation is vital because output from the
speech recognition module will inevitably contain errors
which the machine translation module cannot accept.
This paper proposes a method of generatinq a
Japanese sentence by inferring function words from
content words using valency patterns. This technique is
aimed at the selection of the most appropriate candidate
from a typical phrase lattice that may result from a speech
recognition system.
2 Basic assumptions
In this study the following restrictions relevant to
the interface problem are assumed:
(1) A Japanese sentence usually consists of a certain
number of noun phrases followed by a verb phrase
at the end. The basic unit of speech recognition is
assumed to be a continuously uttered phrase, so
that any input to the machine translation module is
a 'phrase lattice', i.e., a set of phrase candidates
hypothesized by the speech recognition module.
(2) The range of telephone conversation tasks is
restricted to inquiries from a researcher to a clerk
about an international conference concerning the
main topic of the conference, deadlines for paper
submission,
exhibitions,
social
events,
accommodation, payment, cancellation, etc..
restricted number of candidate verbs to be inferred.
This method eliminates possible errors at the
interface between speech recognition and machine
translation (component technologies of an Automatic
Telephone Interpretation system) and selects the most
appropriate candidate from a lattice of typical phrases
output by 1he speech recognition system.
1 Background and problems
An Automatic Telephone Interpretation system is a
facility which enables a person speaking in one language
to communicate readily by telephone with someone
speaking another. Three constituent technologies are
necessary for such a system: speech recognition, machine
translation, and speech synthesis.
Basic research in each of these technologies to
utterance
resultsof
se.p~h
r ~
phrase 1
'genkou-no'
phrase2
'shimekiri-wa'
'genkou' 8 ~ o ' - - 4 I
I'mo' 3 I
L2,oC_
'kentou' 5
~
'shimekiri' 7
IZ _L_I
'shigeki'
I'mo' 3 I
An
J
'desuka" 4
"deshita' 3
"bekika" 1
3
L"_""'_LI
'nentou' 1 [ ~ w o ~
Figure 1
'itsu-desuka '
2
'iku'
"desuka' 4
'deshita' 3
"bekika" 1
'desuka" 4
'deshita' 3
"bekika" 1
exampleofa phraselattice.
291
3 Research goal
3.1 A phrase lattice as the result of speech recognition
Consider a Japanese sentence consisting of t w o
noun phrases and one verb phrase:
'genkou-no shimekiri-wa itsu-desuka'.
('When is the deadline for a manuscript?')
Our method of selecting the correct F-words is
composed of t w o steps: 1) g e n e r a t e a m e a n i n g f u l
sentence by inferring suitable F-words for a given
sequence of J-words, and 2) compare these inferred Fwords with the candidates in the phrase lattice to select
those most appropriate.
Usually a Japanese phrase begins with a jiritsugoword (J-word for short) such as a noun or verb, and ends
with a sequence of fuzokugo-words (E-words for short)
This idea of 'selection-by-generation' distinguishes
this approach from previous ones: Hayes et al. [1] for
English or Niedermair [2] for German. In this paper only
such as postpositional particles or auxiliary verbs. In the
Step 1, which is considered the key step, will be discussed.
above notation, boundaries between J-words and Fwords are explicitly indicated by hyphens, and all F-words
are italicized.
Figure 1 shows an example of a phrase lattice for
this sentence obtained as the result of speech recognition.
Notice that there are candidates for both J-words
and F-words together with a recognition score of the
probability that the word is correct. The problem is to
select the most appropriate candidate from this phrase
lattice.
4 Generating a sentence by inferring F-words
The task domain is restricted to inquiries about an
international conference, and therefore the dialogue is
basically a repetition of simple questions and answers.
This increases the probability of inferring the correct Fwords for each phrase.
4.1 Key information for the inference
The following types of information are used to infer
3.2 Selection-by-generation approach
A t t e n t i o n is focused on candidates for F-words,
assuming t h a t J-words have already been correctly
selected by a suitable method.
The assumption that J-words have been correctly
F-words. The information is described in a lexicon of Jwords.
(1) Semantic features of nouns and valency patterns
First, two types of semantic features are set up for
selected is realistic if the task domain is limited enough to
nouns appearing in the restricted task domain. One is a
general type of semantic feature, independent of the task
a l l o w a high r e c o g n i t i o n rate for J-words and a
domain, such as abstract, action, concrete, human,
knowledge-base, etc. is available for the limited task
location, time, number and diversity. The o t h e r is a
domain. Techniques related to this procedure are now
specific type of semantic feature dependent on the task
being studied. Of the J-words, the predicate a t t h e end of
domain. Table 1 shows examples of such features.
a sentence is less accurately recognized by the speech
Using these semantic features valency patterns of
recognition module than nouns. A method to solve this
the basic predicates necessary in the task domain are
problem will be discussed in the second half of this paper.
In Figure 1, for instance, it is assumed t h a t a
defined. As an example, the predicate 'okuru' ('send' in
English) is given the following valency patterns:
sequence of J-words:
N [ c o n / - t r a ] ' w o ' + V,
'c)enkou'
'shimekiri'
'itsu'
(manuscript')
('deadline')
('when')
has been correctly selected according to the recognition
N [ c o n / - t r a ] ' w a ' + N[Ioc]'ni' + V,
scores. Corresponding to these J-words, there are three
sets of candidates for F-words in the phrase lattice:
'wo'
'too'
'no'
'wa'
' ga"
'desuka'
"deshita '
"bekika'.
Our major concern here isthe subproblem of selecting the
most appropriate one in each of these sets.
N [ c o n / - t r a ] ' w a ' + N [ h u m ] ' n i ' + V,
N [ c o n / - t r a ] ' w a " + N [ t i m / p r o ] ' m a d e n i ' + V,
N [ t i m / p r o ] ' m a d e n i " + N [ c o n / - t r a ] ' w o ' + V,
N[hum]'ni'+
N [ c o n / - t r a ] ' w o ' + V,
N[hum]'ga'+
N[con/-tra]'wo' + V,
N[hum]'ga' + N[con/-tra]'wo' + N[hu m/-pro]'ni' + V,
N[hum/-pro]'wa' + N[con/-tra] ' w o ' + N[h um]'ni'+ V,
This sub-problem is characteristic of the Japanese
etc..
The first valency pattern in this list, for instance, specifies
language. In fact, as easily seen in the above example,
frequently used F-words, specifically those indicating
that the predicate V ('okuru') can take one noun phrase
consisting of a noun with the general semantic feature
grammatical cases such as 'ga', "wo', "ni', etc., are too
short to be recognized correctly. Their recognition scores
'concrete' / specific semantic feature non-'transport' and
F-word ' w e ' .
are much lower than those of J-words. But in Japanese it
In this way the valency patterns summarize the basic
is often possible to infer the meaning of a given sentence
J-word and F-word relationships, and thus give the most
from the sequence of J-words when the task domain is
narrow.
sequence of J-words.
2P)2
essential information for inferring F-words from a given
Table 1
Semantic features
example
gener___a/ specific
abstract
logic
'riron'(theory), 'houhou'(method)
state
'yousu'(state), 'baai'(case)
'language 'nihongo'(Japanese),'eigo'(English)
learning 'bunya'(field), 'senmon'(specialty)
intention 'kyoumi'(interest), 'kibou'(hope)
value 'hitsuyou'(necessity)
sign
'namae'(name)
labor
'youken'(business)
concrete document 'genkou'(manuscript),'youshr(form)
transport 'basu'(bus),'tikatetsu'(subway)
article
'syashin'(photograph)
.___money. 'okane'(money),'kado'(cash card)
pronoun 'kore'(this) ,'dore'(which)
human
h u m a n 'happyousya'(presenter)
pronoun 'clare'(who)
location ~
~kaijyou'(hall),'hoteru'(hotel)
region ~ 'Kyoto'(Kyoto) ,'kaigai'(foreign)
po_.sition 'temoto'(hand) ,'aida'(between)
pronoun 'doko'(where)
time
thee
'jikan'(time), 'ima',(now) 'ato'(after)
pronoun I 'itsu',(when)'nanji'(what time)
number
amount I 'ninzuu'(the number of people)
unlt
I 'en'(yen), 'doru'(dollar)
cost
] 'tourokuryou'(registration fee)
price I 'muryou'(free), 'ikura'(how much)
act
'sanka'(participation), 'yotei'(plan)
diverse 'nani'(what) ,'hoka'(else)
These valency patterns are obtained from the
valency patterns of predicates (obtained from dialogues
collected for the task of inquiries about an international
conference). If necessary, certain modifications such as
omission of the nominative case, modification of the
word sequence or of the F-words, and a d d i t i o n of
interrogative pronouns are carried out. In d i a l o g u e
sentences, the nominative case such as 'watashi' (T) or
°anata' ('you') is seldom used; hence, the nominative case
is usually not included in the valency patterns. To describe
the modification of the two valency patterns for 'okurU'
('send'):
N[tingpro]'madeni" + N[con/-tra]'wo' + V
(a)
N[con/otra]'wa' + N[tim/pro]'madeni' + V
(b)
If the noun N[con/-tra] of the valency pattern (a) becomes
the subjec% the F-word is replaced with 'wa'and the word
sequence is changed, often resulting in the valency
pattern (b). Interrogative pronouns are added to produce
valency patterns specific to i n t e r r o g a t i v e sentences
because a large number of questions occur in this task
domain.
In a ]imited task domain, even individually optional
cases behave in a similar way to the obligatory case for
each predicate. T h e r e f o r e , the o p t i o n a l cases are
described in these valency patterns. When valency
patterns were prepared f o r 65 w o r d s w o r k i n g as
predicates, an average of about 11 valency patterns were
produced for each predicate. Details will discussed in
Chapter 5.
(2) Connection of two nouns by F-word 'no"
it is inferred that nouns which cannot be processed
through valency patterns are likely to be connected with
the F-word 'no' (roughly corresponding to 'of' in English)
in the form 'A no B', where A and B denote nouns. For a
given noun A, the other noun B can also be specified
through the semantic features. For instance, the noun
'kaigr ('conference') can be joined with other nouns as
follows:
'kaigi' + 'no' + N[hum/hum],
'kaigi' + "no' + N[abs],
'kaigi' + 'no' 4. N[tim/tirn],
'kaigi' + 'no' + N[Iodins, pos],
N[abs] + 'no' + 'kaigi',
N[num/amo] + "no' + 'kaigi', etc..
As shown above, whether or not to insert the Fword "no" is automatically determined by presetting
which nouns are to be connected with the F-word "no'.
(3) Syntactic information
Pure syntactic k n o w l e d g e is also useful in this
process. It is known that, in Japanese, no F-word can be
attached to an adverb or a conjunction, and that a verb in
conditional form can be connected with an adjective via
F-words such as'ba'.
In addition, the f o l l o w i n g rules are used, for
example:
Continuative verb form + F-word + verb
--> conjunctive particle 'te',
Conclusive verb form + F-word + verb
-~ conjunctive particle 'to',
Continuative verb form + F-word + adjective
-~ conjunctive particle 'te',
conjunctive particle "temo ',
Attributive verb form + F-word + noun
--7 no F-word,
etc..
4.2 Outline of the process of inferring F-words
Figure 2 (a) illustrates a process of inferring F-words
in a Japanese sentence:
'kaijyou-de kaigi-no yousu-wo
rokuonshi-temo ii,desuka',
(May I record the speech of the conference
a t t h e hall?).
In this case it is assumed that J-words 'kaigi', 'yousu',
'rokuonshi', and 'ii' are correctly recognizable.
The inference proceeds as f o l l o w s : 1) syntactic
information can connect 'rokuonshi' and 'ii' with F-words
'te" or "temo' and 'desuka' to generate the phrases
' r o k u o n s h i - t e ' or ' r o k u o n s h i - t e m o ' and "ii.desuka',
'293
respectively, 2) considering the semantic features of the
first three J-words, and t a k i n g the f o u r t h J-word
'rokuonsuru' as V, the valency pattern:
N[Ioc]'de' + N[act,abs]'wo' + V,
can be applied to them, 3) there are two possible
connections: 'kaigi-no yousu' and 'kaijyou-no kaigi', and
4) considering both 2) and 3) together, sentences:
'kaijyou-(de, no) kaigi-no yousu-wo
N[conl-tra]'ha" + N[timlpro]'madeni' + V.
3) Since the nouns ' g e n k o u ' and 'itsu' cannot be
connected by the F-word "no', 4) the following sentence:
'genkou-wa itsu-madeni okure-ba
yoroshii-desuka',
is finally obtained.
5 An experiment to produce sentences from
rokuonshi-(te, temo) ii-desuka',
can finally be derived.
In a similar way, (b) shows how the f o l l o w i n g
sentence is to be processed:
'genkou-wa itsu-madeni okure.ba
Using the valency patterns obtained from collected
dialogue sentences, we carried out an experiment of
producing sentences. Of the total of 256 interrogative
sentences, 146 were used in d e t e r m i n i n g valency
patterns. The number of verbs was 65 and that of nouns
was 229. In total, 669 valency patterns were prepared
(10.7 patterns for each verb on the average).
In addition to the collected dialogue sentences, we
prepared 70 test questions. For these interrogative
sentences, we carried o u t a s e n t e n c e - p r o d u c i n g
experiment. The results of this experiment are shown in
yoroshii-desuka ',
(By what time may I send the manuscript ?).
Here, 1) 'okure' is combined with 'yoroshii' by the F-words
'ha' and 'desuka', to yield 'okure-ba yoroshii-desuka', 2)
analyzing the semantic features of the nouns 'genkou'
and 'itsu' and the presence of the verb 'okuru', the
following valency pattern is applied:
(a)
valency
patterns
Oriqinal u.tterance
I 'kaijyou-de kaigi-no yousu-wo rokuonshi-temo ii-desuka'
t
1
~ F-words
J
Sequenceof J-wordscorrectly recoqnized
~ to be inferred
...............
Lexiconof J-words
noun
loclins
'/
'i
_'~__:
.......
noun
act
.....
_~_:
'l noun ,
'i abslsta ',
L__~___.:
verb
negativeor
~ c°=t~2u=J~E_,
[
2)Valencypattern - ~ - ~ N [ioi~de'-+
I -N[act~abs]';o'TV
- ~ / - ~ ---
4)Sentenceqeneration
(b)
~F ....=---[ . . . .1)Synt
. . . .a.ct.ic. . . . information
.
j(/
'kaiiyou-Cde..o~ k~igi-.o you.u-wo rokuo,shi-Ire, temo~ ii-,esu~a'
/
Or_~nal utterance
'genkou-wa itsu-madeni okure-ba yoroshii.desuka'-- I
~F-words
~ t o be inferred
Sequenceof J-wordscorrectly recoqnized
L'o e°-u'
.........
noun
m
, ~ con/doc
.,,so. iiill--
,u,e'
r- . . . . . . . . . . . . . . . . . . .
', noun
~
verb
'1 tim/pro.
'~ conditional
,
\
2)Valencypattern. . . . . . ~
I
_~_
. . . . . . . . . . .
=
iiilI
'genkou n o
,,Sente=e
_~.
I semanticfeatures
T--_ . . . . . . .
itsu'
...........
[
,- . . . . . . . . . . . . . .
~. . . . . ~
', Lexiconof J-word=s
adjective
] part of speech
conclusive-~..-~
conjugation
r--
f .....
- LN[c_°_n/d°c]:wa_'_+N[ti.m_]:ma_den!_+V__i l
3)Connection.of..n.oun.s....................... ;
~ r e -
~......
/
....
-~-'.-. ......
ba yoroshii.desuka'l)Syntacti
informat!onj
c
/
'..n.ou-wa ...-ma,e.. o,.=:~a .oro.,,-,.uka'
Figure 2 Process of inferring F-words.
29L
part of speech
conjugation
semanticfeatures
etc.
[_..
I 'r~°V=°"shi'
, / /'e;"°'
(te'/ i'desli'a'
...............
l ]'kaigi
/ rn°'°u'u'
3,Connection of
I nouns
'"aij'°u "° "a'='
/[
I
adjective
conclusive~..
I
1
Table 2. In this experiment, the sequence of J-words of
production of a complete sentence from a given sequence
each test sentence was input, and a complete sentence
of J-words.
including F-words was output. The correct answer rate in
Table 2 is the percentage of all the o u t p u t sentences that
6 Inferring an o m i t t e d verb
were consi~.~tentw i t h the i n p u t sentences. At the first trial,
64.3% c o i r e c t sentences w e r e p r o d u c e d f r o m t h e
prepared valency patterns.
The verb in a given sequence of J-words has an
i m p o r t a n t role in this m e t h o d because it a l l o w s t h e
selection of a correct valency pattern. It w o u l d he difficult
to proceed by this method if the verb is o m i t t e d for some
Table 2 Sen1 ences obtained using valency patterns
Trial
.
.
1
__L_
3
Number - of
4
candidate
5
sentence:; - - ~ -
.
I
.
12
3
.
15
.
4 i
i
16 I
5
16
6
16
1___~4 2___2_0
2~_ z___~d__24
24
0
0___0___ s ,
s --%-_
3
3
3
3
3
3
7
7
7
7
7
7
6
0
2
0
6
0
2
0
6
0
2
0
6
0
2
0
7
0
2
0
5
4
3
0
1
1
1
1
1
1
45
53
59
66
67
70
75.7 I 84.3
94.3
95.7
100
7
8
9
10
Correct sentences
(total)
2
.
.
14
64.3
0
I
---~
.-
81
61
0
6
01
01
0
0
0
0
01
01
61
6
71
0
0
01
4
5
6
11
01
II
31
I
0
I
3
01
01
''
31
0
0
I
3
0
0
0
3
01
01
Ol
ol
Incorrect s~.mtences
(tdtal)
251
17
111
4
3
Ol
6831 690
691
69'41
_3
Valency patterns
(total)
dialogue.
However, in this restricted d o m a i n , nouns w i t h
p a r t i c u l a r s e m a n t i c f e a t u r e s are o f t e n r e l a t e d t o
particular verbs. For example, as shown in Figure 3, in
sentences w h i c h c o n t a i n a noun w i t h the semantic
feature of concrete/document, the noun + F-word ' w o '
tends to be accompanied by the verb 'okuru'('send'),
'kaku'('write') or 'motsu'('have'), and the noun + F-word
'hi'
tends
to
be
accompanied
by
the
verb
'kinyuusuru'('enter').
This suggests the possibility of inferring an omitted
Correct answer
rate (%)
Number
of
candidate
sentences
reason, such as speech recognition failure, or if it were
o r i g i n a l l y o m i t t e d as is o f t e n t h e case in Japanese
669 1 677
verb from the nouns by inversely applying a suitable
valency pattern. In fact, t h e d e f i n i t i o n of a valency
pattern can be generalized as follows:
N[sem Iglsem Is] + N[sem2glsem2s] -t.... + V[v-class],
where V[v-class] denotes a verb belonging to verb class 'vclass'. This valency pattern can be used to infer a verb V[vclass] associated w i t h a set of nouns N [ s e m l g l s e m l s ] ,
N[sem2glsem2s], etc..
This is schematically illustrated in Figure 4. When
there is a verb group A (consisting of verbs which are
inferred when a certain F-word is added to noun A) and a
verb group B (consisting of verbs which are inferred when
The !lpper hatf of Table 2 shows the n u m b e r of
c a n d i d a t e s in t h e trials w h e r e some of t h e o u t p u t
sentences were correct. For example, the number 7 shown
for the first trial in the line corresponding to 5 candidate
a certain F-word is added to noun B),the common area of
these t w o groups indicates the verbs which are inferred
from the valency pattern containing noun A and noun B.
For example, in a sentence which contains a noun
sentences means the number of candidate sentences was
5~ and that 7 of the 70 test sentences were correct ones.
'setsumeisyo'
The l o w e r half of Table 2 shows the number of test
sentences i n ' t r i a l s w h e r e no candidate sentence was
produced or w h e r e none of t h e candidate sentences
'genkou'
~
~ e r b
~ ~
)
produced were correct.
The figures for the second and subsequent trials in
Table 2 show the change' in the correct answer rate when
[concreteldocument
"l )
additional valency patterns w e r e used t o increase the
incidence o f correct sentences. In this experiment, enough
valency patterns were added so th'at the sixth trial always
'tourokuyoushi'
~
~/erb ~
....... \
p r o d u c e d correct sentences.
AI~ elf t h e t e s t sentences u s e d in t h e a b o v e
e x p e r i m e n t w e r e simple i n t e r r o g a t i v e sentences. As
s h o w n in Table 2, tile Use of ~alency patterns allows easy
'
'~shi'
Figure 3 Frequent c o m b i n a t i o n s o f r~ouns and verbs
295
'moushikomiyoushi'
Table 3 Verbs inferred using valency patterns
'temoto'
1
[somonticeoturoA J
/ oncrete,do meot
'WO'
'ni'
2
3
Total
4
1
6
77
29
5
117
2
3
7
2
17
13
3
0
0
27
16
4
3
4
0
0
7
5
1
4
1
0
6
6
0
7
0
0
7
7
1
4
0
0
5
8
o
o
o
2
, . . i
21
--
i
9
0
0
0
0
0
10
2
1
0
0
3
11
12
16
2
2
0
1
0
0
0
0
2
3
1
0
0
0
1
18
1
0
0'
0
1
21
1
0
0
0
1
29
Total
1
30
0
130
0
34
0
5
1
199
=
'aru'
a'
Figure 4 Verbs defined by multiple nouns
with the semantic features of concrete/document and a
noun with the semantic features of location/position,
verbs such as 'aru'('be') and 'motsu'('have') tend to be
I
A: Number of verbs inferred
B : Number of nouns in valency patterns
selected, and F-words specific to these verbs are chosen.
Table 3 shows the n u m b e r of verbs which are
inferred from a given sequence of nouns using the
results indicate that, in a restricted task domain, the
valency patterns described in Chapter 5. Since valency
patterns were prepared for 65 verbs, the verbs are
semantic features of the preceding nouns and valency
inferred -from these 65.
The columns in Table 3 Show the number of verbs
inferred. The lines in the same table show the number of
verbs to be inferred.
nouns in the valency patterns. For example, when the
number of inferred verbs is 5, there are 6 valency patterns
where 5 verbs are inferred; and of these 6 patterns 1 has
one noun, 4 have 2 nouns, 1 has 3 nouns and none have 4
nouns.
In counting the number of inferred verbs, only verbs
having a valency pattern consistent with given valency
patterns are counted. When the noun of a valency
pattern bears a specification as to the upper-level general
semantic features but no specification as to the lowertask~dependent semantic features, the verbs of the
valency patterns bearing a specification as t o t h e lowerevel semantic features are counted. Conversely, for
valency patterns where the lower-level semantic feature
is specified, the verbs bearing no specification as to the
Iower.-levei semantic features are n o t c o u n t e d . For
example, in the f o l l o w i n g t w o valency patterns, the #erbs
Vl and V2 are inferred from the pattern (a), while only
the verb V2 is inferred from the pattern (b).
patterns allow a fairly restricted number of candidate
7 Conclusions
As a first step toward a better interface between
speech recognition and machine translation, a method
which is particularly useful for Japanese sentences was
proposed to infer F-words for a given sequence of Jwords.
In a restricted task domaia., the most appropriate Fword can be inferred from a given sequence of J-words if
the task-dependent semantic features of nouns are preset
and the information of valency patterns is utilized.
In addition, the results of this study suggest that
correct verbs can be inferred from valency patterns.
The authors are now evaluating the effectiveness of
the procedures proposed in this paper by applying them
to actual results of speech recognition.
Acknowledgment
The authors are deeply grateful to/Dr. K~rematsu,the president of
/ATR Interpreting Telephony Research Eaboratories, and all the
members of ATR Interpreting Te/lephor~y Research Laboratories for
their constant help and encouragement:
N[con] + N[Ioc] + V l
(a)
N[con/doc] + N[Ioc] + V2
(b)
As can be seen in Table 3, only one verb was inferred
in more than 50% of the valency patterns. Irl 90% of the
remaining valency patterns where multiple verbs were
in~erred, the number c~f verbs i~fe~rred was 6 o~ le~s. ~ h e s e
29~
References
[1] Hayes, P.J.et al., "Parsing Spoken Language; A Semantic
[2]
Caseframe Approach", EOLING,86
Niedermair, G., "Divided a~d Valency-Oriented Parsing in
SpeechUnderstandir)g",COLING86.